Systematic review of the clinical effectiveness and cost ... - CiteSeerX

“Lap. surg. syst. review economic eval.”

This article was originally published in Colorectal Disease 2008;10(9): 85968 DOI: 10.1111/j.1463-1318.2008.01701.x and is available from URL: http://www3.interscience.wiley.com/journal/121494305/issue

Systematic review of economic evaluations of laparoscopic surgery for colorectal cancer

Authors: Rodolfo A. Hernández1, MSc.; Robyn M. de Verteuil1, MSc.; Cynthia M. Fraser2, MA.; Luke D. Vale1,2, PhD; On behalf of the Aberdeen Health Technology Assessment Group.

1. Health Economics Research Unit. Institute of Applied Health Sciences, College of Life Sciences and Medicine, University of Aberdeen. Aberdeen, Scotland, UK. 2. Health Services Research Unit. Institute of Applied Health Sciences, College of Life Sciences and Medicine, University of Aberdeen. Aberdeen, Scotland, UK.

Acknowledgements: The paper was developed from a Technology Assessment Review conducted on behalf of the National Institute for Health and Clinical Excellence (NICE) and was funded by the Department of Health on a grant administered by the National Coordinating Centre for Health Technology

1


Assessment.

The authors are grateful for the comments from independent

reviewers of the Technology Assessment Review on which this paper is based. The Health Economics Research Unit and the Health Services Research Unit are core funded by the Chief Scientist Office of the Scottish Government Health Directorates.

The views expressed here are those of the authors and not

necessarily those of the funding bodies.

Correspondence to: Rodolfo Hernández, Research Fellow, Health Economics Research Unit, University of Aberdeen, Polwarth Building, Foresterhill, Aberdeen, AB25 2ZD. e-mail: [email protected] Tel.: +44-1224-558992 Fax: +44-1224-550926

Word count abstract: 248 Word count main text: 4287

2


ABSTRACT Background Colorectal cancer is common and the standard surgical treatment is open resection (OS) but laparoscopic surgery (LS) maybe an alternative. In 2000, a Health Technology Assessment (HTA) review found little evidence on costs and cost-effectiveness comparing the two methods. The evidence base has since expanded and this study systematically reviews the economic evaluations on the subject published since 2000.

Methods Systematic review of studies reporting costs and outcomes of LS vs. OS for colorectal cancer. National Health Service Economic Evaluation Database (NHS EED) methods for abstract writing were followed. Studies were summarised and incremental cost-effectiveness ratios (ICER) for common outcomes calculated.

Results Five studies met the inclusion criteria. LS generally had higher health care costs. Most studies reported longer operational time and shorter length of stay and similar long-term outcomes with LS vs. OS.

Only one outcome,

complications, was common across all studies but results lacked consistency (e.g. in two studies open surgery was less costly but more effective, in another laparoscopic surgery was less costly but more effective, and in the further two

3


laparoscopic surgery could potentially be cost effective depending on the decision-makers willingness to pay for the health gain).

Conclusion The evidence on cost-effectiveness is not consistent. LS was generally more costly than OS. However, the effectiveness data used in individual economic evaluation was imprecise and unreliable compared with data from systematic reviews of effectiveness. Nevertheless, short-term benefits of LS (e.g. shorter recovery) may make LS appear less costly when productivity gains are considered.

KEYWORDS Economic evaluation, cost effectiveness, systematic review, laparoscopic surgery, colorectal cancer

4


1

Introduction

Colorectal cancer is one of the most common cancers in the Western world and is the second most common malignancy in England and Wales in terms both of incidence and mortality. Approximately 36,000 new cases were diagnosed in 2002 and 17,000 people died from colorectal cancer in the same year. About 80% of all patients diagnosed with colorectal cancer (including some with advanced disease) undergo surgery.[1]

Open resection is currently the standard method for surgical removal of primary colorectal tumours. Minimally invasive approaches to treat colorectal diseases were developed to take advantage of the benefits observed in laparoscopic procedures elsewhere in the gastrointestinal tract.[2] Laparoscopic surgery, includes total laparoscopic, laparoscopic assisted and hand assisted procedures. In practical terms a totally laparoscopic and laparoscopicallyassisted procedure are considered comparable because of the size of incisions involved. In the remainder of this paper laparoscopic and laparoscopicallyassisted surgery are referred to as laparoscopic surgery.

In hand assisted

laparoscopic surgery (HALS), the surgeon inserts a hand into the abdomen while pneumoperitoneum is maintained. Some surgeons find this easier than laparoscopic

surgery

particularly

in

the

transitional

phase

between

conventional and laparoscopic surgery. Advantages claimed for placing the hand in the abdomen include tactile feedback, the ability to palpate, blunt dissection, organ retraction, control of bleeding, and rapid organ removal.[3-5] 5


Laparoscopic surgery is less invasive and hence may lead to more rapid recovery from the operation but the potential impact on cure rates is unclear. The major concerns are that tumour recurrence might occur at port sites and that clearance of the tumour may be less complete than during open surgery. It has also been suggested that the reduced trauma to tissues may lower disruption to the immune system and hence reduce the risk of recurrence.[6] Additionally, it has been argued that there are disadvantages of laparoscopic surgery relating to the longer length of the operation, the cost of materials, and the effect of surgeon experience on patient outcomes.

Vardulaki and colleagues[7] conducted a review of studies comparing the costs of the laparoscopic surgery and open surgery in 2000, (there were no economic evaluations available at this time), as part of the health technology assessment review (TAR) carried out by the National Institute for Clinical Excellence. Eight studies met their inclusion criteria and they reported that there was no significant difference in costs between the two procedures in most studies. Differences in cost were driven mainly by the estimates of cost of hospitalisation and time in theatre which varied between studies. Moreover, Vardulaki and colleagues[7] conducted a costing exercise but did not combine data on costs with data on effectiveness in an economic evaluation to assess costeffectiveness. In addition to this, the evidence base (for effectiveness and cost)

was very limited as no studies reported long-term outcomes and economic

6


evidence was of limited quality. Therefore, in terms of efficiency no robust conclusions could be drawn from the HTA 2000 report.[7]

Since 2000 the evidence base has expanded and experience with laparoscopic technique has increased.[1] The present study is a systematic review of the economic evaluations conducted as part of a health technology assessment on laparoscopic surgery for colorectal cancer conducted on behalf of the UK’s National Institute for Health and Clinical Excellence (NICE). Amongst other reasons, systematic reviews are important as they can establish whether scientific findings are consistent and can be generalisable to other populations and/or settings. Furthermore, explicit methods used in systematic reviews limit bias and improve reliability and accuracy of conclusions.

The paper reports a systematic review of the economic evaluations of laparoscopic and/or hand assisted laparoscopic techniques compared to open surgery for the treatment of colorectal cancer.

7


2

Methods

2.1

Search Strategies

Studies that reported both costs and outcomes of laparoscopic and/or HALS techniques compared to open surgery for the treatment of colorectal cancer were sought from a systematic review of the literature.

No language

restrictions were imposed but as this review is an update of an earlier review conducted in 2000, the searching was limited to studies available between 20002005.

Databases searched were Medline (2000 – May Week 2 2005), Embase (2000 Week 21 2005), Medline Extra (23rd May 2005), Science Citation Index (2000 – 27th May 2005), National Health Service Economic Evaluations Database -NHS EED- (May 2005), Health Technology Assessment (HTA) Database (May 2005), Health Management Information Consortium (2000 – May 2005) and Journals @ Ovid Full Text (2000 – July 2005 for selected surgical journals). In addition, recent conference proceedings and reference lists of all included studies were scanned to identify additional potentially relevant studies. Other sources of information consulted included: references in relevant articles; selected experts in the field; references of consultees’ submissions to NICE. Full details of the search strategies used are available from the authors.

8


2.2

Inclusion and exclusion criteria

To be included, studies had to compare, in terms of costs and outcomes, strategies involving laparoscopic and/or HALS compared to open surgery for treatment of colorectal cancer. Studies were included even if they made no formal attempt to relate cost to outcome data in a cost-effectiveness or costutility analysis.

One reviewer assessed all abstracts for relevance and full

papers were obtained for those that appeared potentially relevant.

2.3

Data extraction strategy

The following data were extracted for each included primary study using the framework provided for abstracts prepared for the NHS Economic Evaluation Database.[8]

a) Study identification information: author and year, interventions studied, type of economic evaluation, country of origin and currency reported; b) the intervention, study design and main outcomes: fuller description of treatment, numbers receiving or randomised to each intervention, outcomes studied; c) Sources of data: effectiveness data, mortality and comorbidity (if measured), cost data, quality of life (if measured); d) methods and study perspective; e) results: costs, benefits, incremental cost-effectiveness/utility ratio (ICER), sensitivity analyses; f) additional comments relating to the design and reporting of the economic evaluation.

9


No primary outcome was specified although all outcomes were prespecified in a protocol developed before the study started. Ideally all these measures might be synthesised into a single unitary measure such as quality adjusted life years (QALYs). In absence of such outcomes, data were abstracted on the following outcomes:

mortality,

survival;

disease

free

survival;

recurrences

and

complications (this latter outcome was the only one common across all studies).

2.4

Data synthesis

In economic evaluations, as well as in other research areas, is very important to know which methods the authors used to develop their calculations and analyses, as the results could eventually vary widely according to different methodologies used.[9] For instance, cost categories should be reported and it is desirable that health care resource used as well. Moreover, no matter what methodology is used, any economic evaluation is subject to uncertainty. The uncertainty (e.g. the type and sequence of relevant events considered, variability in data, etc.), can be addressed by conducting different types of sensitivity analyses.[10] Sensitivity analyses will tell the reader how reliable the results are.

One economist assessed included studies using the NHS-EED guidelines for reviewers.[8] This guidelines address all the important issues that should be reported when conducting an economic evaluation in health care. No attempt was made to synthesize quantitatively the primary studies that were identified. Data from all included studies were instead summarised and appraised in 10


order to identify common results, variations and weaknesses between studies. If a study did not report incremental cost effectiveness ratios (ICERs) but provided sufficient data then, where possible, the data were reanalysed to provide estimates of ICERs.

These ICERs are presented for the following

clinical outcomes: mortality, survival; disease free survival; recurrences and complications.

3

Results

3.1

Number of studies identified

As a result of the literature searches, 392 study hits after deduplication were screened (Figure 1). From these, a total of 41 studies were selected for full assessment. Of these 41 studies, three studies[11-13] met the inclusion criteria. Two additional unpublished papers (that have since been published) also meeting the inclusion criteria were obtained from experts in the field.[14,15] A further study that compared laparoscopic surgery against HALS and, as a consequence did not meet the inclusion criteria, was also identified.[16]

3.2

Study identification and key elements

Two studies compared laparoscopic colon resection with open colon resection in the treatment of colon cancer,[11,13] one of which focused on right hemicolectomy.[13] A further study compared laparoscopic-assisted with conventional open resection for rectosigmoid carcinoma,[12] and two compared

11


laparoscopic versus open resection for colorectal cancer.[14,15] One of these was in the context of an enhanced recovery program.[15]

Five studies were classified as cost-consequence analyses. That is, costs were compared with various different measures of effectiveness. Two were based on single centre RCTs[12,15], and two were based on multicentre RCTs.[11,14] The fifth study was based on a single centre cohort-matched study conducted in China (Table 1).[13]

Two studies considered costs from a societal

perspective[11,15] while the others adopted a hospital perspective (Table 1).[12,13] Franks and colleagues described their study as a cost analysis but included data on outcomes sufficient to calculate ICERs.[14]

The study by Franks and colleagues represented a preliminary analysis conducted on a subset of patients from the CLASICC trial[17] who had agreed to be included in the economic analysis. The dates for data collection were not reported. The Swedish study collected data from January 1999 to May 2002;[11] the study by King and colleagues from January 2002 to March 2004,[15] the study by Leung and colleagues, conducted in Hong Kong, collected data from September 1993 to October 2002,[12] and the Chinese study from September 2002 to February 2003.[13] In all five studies costs were estimated prospectively from the same sample as that used for collecting the effectiveness data.[11-15]

12


3.3

Patient group, study sample and study design

The sample sizes in four of the five studies were modest (Table 1). In the cohort matched study, 30 patients with colon cancer underwent laparoscopic right hemicolectomy surgery and were matched with 34 patients who received open right hemicolectomy surgery.[13] Patients for the open surgery group in this study were matched for gender, age, Dukes’ staging, tumour site, previous abdominal operation and extent of resection. 34 patients were randomly selected from 87 patients who underwent open surgery during the same period.

The analysis in all studies was conducted on an intention to treat basis, however, the follow-up period as well as the outcome measures varied considerably between studies (Table 1).

3.4

Methods of economic analysis

The four trial based papers[11,12,14,15] presented details on which items were included in the cost calculations, but no details were reported in the nonrandomised study.[13] Such information is useful as data on resource use can help readers judge the applicability of the study to their setting. Relatively good details of unit costs were presented in the Swedish and UK studies[11,14,15] while no unit costs were reported in the other two studies.[12,13] In economics, costs occurring into the future are given less weight than costs occurring now, i.e. they are discounted. Discounting was performed only in the Swedish study while it was actually relevant in all studies with a follow-up greater than 12 months. Indirect costs were calculated 13


in three of the studies using the human capital approach (time off paid work)[11,14,15] Three papers did not use any summary measure of health benefits[12,13,15] and left the results disaggregated. One study focused primarily on costs.[14] In the Janson and colleagues study, the mean cost per reoperated patient for each arm of the trial was presented (although it is not reported in this paper).[11]

As previously stated, uncertainty is pervasive in every economic evaluation. Therefore, authors should allow for this by conducting different types of sensitivity analyses. One-way sensitivity analysis was performed in three studies.[11,14,15] Changes in perioperative, equipment, recovery, ICU and hospital costs were considered in the study by Franks and colleagues. They also considered a subgroup analysis by location of cancer (colon or rectum).[14] Cost per minute for the operating room, anaesthesia and recovery room time were explored in the Swedish study[11] while duration of in-patient stay and the consumption of community resources after discharge were explored in the Study by King and colleagues.[15]

3.5

Cost effectiveness results from the included studies

The results of the included studies are summarised in Table 2. In Franks and colleagues, total costs, including productivity costs, for both the base-case and both subgroups were not significantly different between the laparoscopic and open groups, although the confidence interval was very wide (mean additional cost of laparoscopic surgery was £268, 95% CI: –690 to 1460). Productivity costs 14


were not a major determinant of this additional cost although hospitalisation costs (less for laparoscopic surgery) and costs for re-operations and other complications (greater for laparoscopic surgery) were. Although there was no evidence of a statistically significant difference in clinical outcomes, confidence intervals would be sufficiently wide for clinically and economically important differences to exist.[14]

In Janson and colleagues total costs, including productivity loss, were not significantly different between the laparoscopic surgery and open surgery. However, total costs, excluding productivity losses (that is cost to the healthcare system), were significantly higher for the laparoscopic surgery compared with open surgery (€9474 vs. €7235; P=0.018), as were the costs related to the first admission, and the costs of primary surgery.[11]

In King and colleagues the results reflected the increased duration of laparoscopic procedures and also the increased use of disposable equipment in theatre. However, in their analysis, King and colleagues found that these costs were more than offset by lower post-operative costs such as re-operations, and productivity cost savings resulting from the earlier return to usual activities.[15]

Similarly, the health service costs from Leung and colleagues were also higher for laparoscopic surgery than for open surgery and this difference, as with the other two RCT-based analyses, was statistically significant (P